Marine Biology (2002) 140: 425±434 DOI 10.1007/s00227-001-0723-3
S. Pertola á M. Koski á M. Viitasalo
Stoichiometry of mesozooplankton in N- and P-limited areas of the Baltic Sea
Received: 9 April 2001 / Accepted: 20 September 2001 / Published online: 9 November 2001 Ó Springer-Verlag 2001
Abstract The Baltic Sea is a very suitable site for stoichiometric studies, since its subbasins dier in their concentration of elemental components, and primary production can therefore be either nitrogen or phosphorus limited. To reveal if the nutrient limitation of mesozooplankton mirrors that of the primary producers, carbon, nitrogen and phosphorus content of both seston and grazers (Acartia sp., Centropages hamatus, Daphnia cristata, Eurytemora anis, Limnocalanus macrurus, Temora longicornis) were measured in midsummer in the Baltic proper, the Gulf of Finland and the Gulf of Bothnia. The mineral ratios of the dierent taxa were equal, apart from L. macrurus with notably higher C:P and N:P ratios. Molar C:N ratios were relatively stable (5.1±6.3), whereas C:P and N:P ratios ¯uctuated more (41±144 and 6.6±24). However, zooplankton elemental composition and limitation did not depend on the limiting nutrient of the phytoplankton, the seston mineral ratio or the sea area. Both the seston±zooplankton elemental imbalance and the food threshold ratio indicated
Communicated by L. Hagerman, Helsingùr S. Pertola (&) Finnish Institute of Marine Research, P.O. Box 33, 00931 Helsinki, Finland E-mail: sari.pertola@®mr.® Fax: +358-9-61394494 M. Koski TvaÈrminne Zoological Station, 10900 Hanko, Finland M. Viitasalo Department of Ecology and Systematics, Division of Hydrobiology, P.O. Box 17, 00014 University of Helsinki, Finland Present address: M. Koski Netherlands Institute of Sea Research, P.O. Box 59, 1790 AB Den Burg, Texel, The Netherlands Present address: M. Viitasalo Finnish Institute of Marine Research, P.O. Box 33, 00931 Helsinki, Finland
phosphorus limitation of most of the grazers. While L. macrurus may be C or N limited, the possible P de®ciency of the other studied taxa suggests that the Baltic Sea zooplankton may act as a potential phosphorus sink, as the freshwater secondary producers do.
Introduction The stoichiometric theory derives from the assumption that mineral composition of the zooplankton remains fairly constant even if the elemental ratio of the food varies (Sterner and Hessen 1994). Accordingly, stable mineral ratios have been observed frequently (Sterner 1990; Andersen and Hessen 1991; Pagano and Lucien 1993; Urabe 1993; Elser and Urabe 1999; Koski 1999), but also more variable ratios occur (Cataletto and Umani 1994; Gismervik 1997a; Koski and Viitasalo, unpublished data). While a given zooplankton species can have a roughly constant C:N:P ratio, this ratio can clearly vary from taxon to taxon. For instance, the C:N and C:P ratio of the large carnivorous zooplankton may be low compared to its food (Malej et al. 1993), and the mesozooplankton (200±2000 lm) can have a remarkedly lower C:P ratio than the smaller microzooplankton (20±200 lm) (Gismervik 1997b; Gulati and DeMott 1997). Dierences also occur within mesozooplankton, as cladocerans typically are phosphorus rich (Andersen and Hessen 1991; Gismervik 1997a), and copepods usually contain a higher proportion of nitrogen (Elser and Hasset 1994; Gismervik 1997a). Since the C:N and C:P ratios of the seston are usually higher than those of zooplankton (Urabe and Watanabe 1992; Brett et al. 2000), cladocerans with a high phosphorus demand should face P limitation (Sterner 1993; Hasset et al. 1997; Urabe et al. 1997), whereas copepods with a higher N:P ratio should be more vulnerable to N limitation (Sterner and Hessen 1994). This is supported by several laboratory studies (Checkley 1980; Kiùrboe 1989; Sterner 1993; Hasset et al. 1997; DeMott et al.
426
1998; Koski et al. 1998) and models (Touratier et al. 1999). On the other hand, it has been pointed out that, due to their low carbon net production eciency, marine copepods can be limited by the availability of carbon rather than nitrogen (Anderson and Hessen 1994). Nutrient de®ciency can be estimated from the critical food threshold ratio, Q*c-e=KcQz-e, where Kc is the gross growth eciency (in carbon), Qz-e the elemental ratio of the consumer, and Q*c-e the critical threshold ratio, assuming 100% assimilation eciency of the mineral elements (Urabe and Watanabe 1992). Because excretion of the limiting nutrient hardly ceases during nutrient limitation, the realised threshold ratio is likely to be lower (Urabe and Watanabe 1992; DeMott et al. 1998; Elser and Urabe 1999). Lower threshold ratios of the zooplankton compared to the elemental ratio of the food (C:N, C:P) indicate nutrient limitation (N or P), while higher threshold ratios may mean carbon limitation. If carbon is abundantly available, the relative importance of the nutrients can also be estimated from the dierence N:Pseston±N:Pzooplankton, where the positive elemental imbalance implies phosphorus limitation, and the negative imbalance indicates nitrogen de®ciency (Elser and Hasset 1994). Furthermore, to maintain homeostasis of body nutrients, zooplankton assimilates the limiting nutrient more eciently, while the nutrients in surplus are released at a higher rate (Sterner et al. 1992; Gulati and DeMott 1997). The N:P of excreted nutrients is primarily determined by the dierence between the food N:P and the N:P ratio of the grazer (Elser and Urabe 1999). Consequently, in a phosphorus-limited environment with a high food N:P-ratio, P-limited grazers (such as daphnids) tend to exaggerate the P limitation, whereas N-limited grazers (copepods) alleviate it (Urabe 1993). In areas where and during periods when zooplankton grazers are abundant, their nutrient recycling can modify the nutritional status of the water and hence aect phytoplankton production and species composition (Urabe 1993). The brackish-water Baltic Sea is a very suitable arena for stoichiometric studies, since its subbasins dier in their elemental characteristics. Phytoplankton production in the Baltic proper and the Gulf of Finland is generally nitrogen limited, whereas in the Bothnian Bay and in the coastal areas of the Bothnian Sea, primary production is mostly phosphorus limited (Wul et al. 1986; GraneÂli et al. 1990). The mineral limitation of the zooplankton in the Baltic Sea is poorly known. However, recent studies in the Baltic proper and the northern Baltic Sea suggest nitrogen de®ciency of planktonic crustaceans during high seston C:N ratios in spring and early summer (Koski 1999; Walve and Larsson 1999). The present study shows elemental composition of the dominant copepod and one cladoceran species in the dierent parts of the Baltic Sea. We aimed to see whether the mineral limitation of the phytoplankton is re¯ected in that of the typical zooplankton in the different areas, and to reveal the possible role of these
secondary producers in nutrient recycling. We hypothesise that, like primary producers, also zooplankton are P limited in the Bothnian Bay and the Bothnian Sea, and N limited in the Gulf of Finland and the Baltic proper. Thus, depending on the sea area, zooplankton would act as a sink either for phosphorus or nitrogen.
Materials and methods The samples were collected at eight open sea stations during 29 June to 8 July 1999. The study area ranged from the Central Baltic proper (57°15¢N; 20°05¢E) to the Bothnian Bay in the north (65°23¢N; 23°28¢E) and the Gulf of Finland in the east (60°04¢N; 26°21¢E) (Fig. 1). The sampling stations were selected to re¯ect the varying nutritional conditions in the Baltic Sea, so that both nitrogen-limited (the Central Baltic proper and the Gulf of Finland) and phosphorus-limited (especially the Bothnian Bay) areas were included. Surface water salinity varied from 2 to 3 PSU in the northern Gulf of Bothnia (stations F2, BO3) to ca. 5 PSU in the Bothnian Sea (SR5) and to 6±7 PSU at the entrance to the Gulf of Bothnia (IU5), in the Gulf of Finland (LL3A, LL12) and in the Baltic proper (LL23, BY15) (Finnish Institute of Marine Research, unpublished data). A more detailed description of the study area is presented, for instance, in Voipio (1981) and HELCOM (1996). Samples were collected at 0800±0900 hours (BO3, SR5, LL12, LL23, BY15) or in the early evening between 1600 and 1900 hours (F2, IU5, LL3A). Zooplankton was sampled with a 100 lm plankton net (diameter 57 cm) with a single haul from ca. 40 m to the surface. At the stations where the surface water layer was dominated by cyanobacteria aggregates (LL3A, LL12, LL23, BY15) the net was closed at 10±15 m depth. The animals were placed in a 30 l container with water from 5 m depth and stored at 4.3°C until treated. The taxa collected diered partly between the stations (Table 1). For the seston mineral composition and chlorophyll a (chl a) analyses ca. 10 l of surface water (0±1 m) was sampled at every station. From this water, fractionated subsamples
Fig. 1 Map of the Baltic Sea showing the sampling stations. Depth at the stations: F2: 89 m, BO3: 109 m, SR5: 128 m, IU5: 89 m, LL3A: 69 m, LL12: 83 m, LL23: 457 m, BY15: 242 m
427 Table 1 Zooplankton species collected at the dierent sampling stations during the cruise of R.V. ``Aranda''
Species Acartia sp. Centropages hamatus Daphnia cristata Eurytemora anis Limnocalanus macrurus Temora longicornis
F2
BO3
SR5
IU5
LL3A
LL12
LL23
BY15
+
+
+
+
+ +
+ +
+
+
+
+
+
+
+ + +
+
(<20 lm and total) were taken for both seston and chl a measurements. For the C:N analysis, adult female copepods and adult cladocerans were sorted out individually under a binocular microscope. Each animal was washed two times with Milli-Q water, placed into tin cups (8±10 individuals per sample) and dried at 40± 60°C for at least 24 h. The carbon and nitrogen contents were analysed with a stable isotope analyser (RoboPrep-TracerMass ANCA-MS) (Barrie and Lemley 1989). For the P analysis, 8± 10 individuals per sample were picked, washed in Milli-Q water, placed on combusted GF/F ®lters, and dried at ca. 20°C. Then 2 ml of 0.017 M MgSO4 was added, and the samples were dried at 95°C for 24 h and analysed with the molybdate method (SoloÂrzano and Sharp 1980). For each species and sampling station, two to ®ve replicates were analysed. To measure the seston particulate carbon (POC), nitrogen (PON) and phosphorus (POP) contents and chl a concentration, two replicate water samples of 100 ml were ®ltered on combusted GF/F ®lters. Two size classes (<20 lm and total) were measured for each sample. Filters for the POC, PON and POP analyses were treated similarly to those of the zooplankton samples. Blanks were prepared for quality control. The blanks were treated similar to the samples, but pure combusted GF/F ®lters (for C and N analyses) and Milli-Q water (for P analysis) replaced the sampled material. Chlorophyll a was extracted from ®lters in 10 ml 96% ethanol, sonicated and incubated in darkness for 18±24 h. The samples were then analysed with a Perkin-Elmer LS-2B ®lter ¯uorimeter. Water nutrients were analysed using a Skalar 5100 autoanalyser applying accredited methods, with detection limits of 0.05 lmol l±1 for both nitrate and phosphate (Grassho et al. 1983). Carbon and nitrogen were analysed using the same replicate, and C:N ratios were calculated for each replicate. Since phosphorus was measured separately, in C:P and N:P ratios each C and N replicate was divided by the mean of the corresponding phosphorus replicates. Dierences in zooplankton mineral ratios were tested with one-way analyses of variance (ANOVA). To reveal intraspeci®c and areal dierences or interspeci®c dierences, station or species was used as the classifying factor, respectively. Pairwise comparisons were made with Tukey's a posteriori HSD-test. If the species was collected only at two sampling sites, t-test was used instead of ANOVA in intraspeci®c and areal comparisons. These analyses were conducted using SAS procedures. To compare the mineral composition of the zooplankton and the seston, non-parametric Spearman correlations were calculated using the SYSTAT 7.0 statistical package. In the correlation matrix all zooplankton replicates were compared with the mean seston value at the respective site.
+
SR5 in the Bothnian Sea and at LL23 in the Baltic proper (0.00 lmol PO4 l±1). Phosphate concentration at 1±10 m depth was below or at the detection limit at the other sites as well; in these cases, extremely low concentrations were still measured (0.01±0.05 lmol PO4 l±1). On the other hand, nitrate was always more abundant (0.09±6.30 lmol NO3 l±1). Phosphate concentrations below the detection limit may not be truly comparable, but since NO3 increased notably towards the northernmost sites, water NO3:PO4 ratios indicated stronger phosphorus de®ciency in the north. Molar NO3:PO4 ratio at 1±10 m depth was remarkably high at F2 in the Bothnian Bay (360±630) and at BO3 in the Quark (100±470). At the rest of the sites, the water NO3:PO4 ratio was near the Red®eld ratio at 1 m depth; 15, 17, 14 and 12 at IU5, LL3A, LL12 and BY15, respectively. At 5±10 m depth these four sites had more variable NO3:PO4 ratios (4±27). Therefore, phosphate limitation of the primary production seemed likely at stations F2, BO3, SR5 and LL23, whereas at the other sites (IU5, LL3A, LL12, BY15) nutrient limitation of the primary producers was not obvious. Seston The seston chlorophyll a, particulate organic carbon and particulate organic nutrient concentrations were lower and the seston C:N and C:P ratios were higher in the Gulf of Bothnia (F2, BO3, SR5, IU5) than elsewhere in the sampling area. Further, a larger proportion of the seston pool in the Gulf of Bothnia consisted of the <20 lm size fraction compared to the other sampling sites (Table 2). The proportions of the heterotrophs and other organic particles versus the autotrophs (total POC:total chl a) varied. The high seston POC:chl a ratio at SR5 (the Bothnian Sea) in both size fractions indicated that much of the local seston consisted of heterotrophs and detritus, whereas especially at LL12 in the Gulf of Finland the share of autotrophs was higher (Table 2).
Results Water phosphate and nitrate
Intraspeci®c dierences and areal comparison
The uppermost water layer (1±10 m), where most of the primary producers occurred, was phosphate depleted at
The C:N ratio of most of the zooplankton species was relatively stable, and generally no statistically signi®cant
428 Table 2 Chlorophyll a (chl a), particulate organic carbon (POC), nitrogen (PON) and phosphorus (POP) concentrations (lg l±1), the elemental ratios (lM l±1:lM l±1), and POC:chl a ratio (lg l±1:lg l±1) Station
POC
<20 lm size fraction F2 520.814.8 BO3 476.2210 SR5 443.88.7 IU5 556.650.5 LL3A 645.163.3 LL12 476.36.6 LL23 655.94.2 BY15 699.487.5 Total size fraction F2 551.716.9 BO3 541.14.6 SR5 458.0104 IU5 562.734.9 LL3A 966.117.4 LL12 762.796.3 LL23 880.160.1 BY15 812.530.9
PON
POP
of seston in <20 lm and total size fractions (meanSD). Where no SD is given, the value is from one sample only
POC:PON
POC:POP
PON:POP
Chl a
POC:chl a
41.42.6 33.223.1 35.10.1 46.612.6 60.48.6 43.61.3 62.30.5 60.318.5
4.40.07 4.50.07 3.70.1 5.10.1 8.90.0 6.00.1 6.90.8 7.60.2
14.70.5 18.75.6 14.80.2 14.32.6 12.50.6 12.70.2 12.30.03 13.92.6
308.913.8 275.0117.0 309.45.8 281.933.3 186.918.3 204.87.6 248.629.8 239.436.6
21.11.6 16.411.2 21.00.7 20.36.0 15.02.1 16.10.8 20.32.5 17.75.9
2.50.1 2.50.0 1.10.07 2.0 2.60.07 3.00.1 2.60.2 2.80.4
208.711.8 190.50.0 423.628.5 278.3 253.17.0 158.47.5 258.121.5 252.738.3
45.51.3 44.91.1 40.321.1 45.99.3 130.50.2 102.614.2 108.410.9 83.91.4
4.30.6 4.60.07 5.20.1 5.80.1 10.90.4 8.60.6 9.50.3 8.50.1
14.20.02 14.10.5 14.44.6 14.52.1 8.60.1 8.70.1 9.50.3 11.30.6
337.840.3 306.77.4 226.545.2 252.312.6 229.73.3 228.213.8 238.89.2 246.55.3
23.92.9 21.80.2 17.08.5 17.63.4 26.60.8 26.31.9 25.21.8 21.80.7
3.7 3.0 1.0 2.8 7.6 7.2 7.5 4.7
149.1 180.4 458.0 201.0 127.1 105.9 117.3 172.9
intraspeci®c dierences were observed (ANOVA; P>0.05) (Tables 3, 4). An exception was Acartia sp. (ANOVA; F5=6.5, P<0.003), the C:N ratio of which was highest in the Gulf of Finland (Tukey's HSD-test; P<0.05). The same tendency, even if not signi®cant, appeared intraspeci®cally also among other taxa. Therefore, the zooplankton C:N ratio was signi®cantly higher in the Gulf of Finland than in the Baltic proper (Tukey's HSD-test; P<0.05). As expected, the zooplankton C:P and N:P ratios varied somewhat more than the C:N ratio (Gismervik 1997a). There were signi®cant intraspeci®c dierences in the C:P ratio of Acartia sp., Eurytemora anis and Limnocalanus macrurus (ANOVA; F5=5.7, P<0.006, F4=8.1, P<0.004 and F4=±3.6, P<0.01, respectively),
and in the N:P ratio of Acartia sp. and E. anis (ANOVA; F5=7.6, P<0.002 and F4=8.9, P<0.003, respectively; Tables 3, 4). In contrast, the C:P and N:P ratios of Centropages hamatus and Temora longicornis were relatively stable (ANOVA and t-test; P>0.05). Because of the relatively high intraspeci®c variation, no unambiguous areal trends occurred in the zooplankton C:P and N:P (Table 3). However, mainly because of the high elemental ratios of L. macrurus, the zooplankton C:P and N:P ratios were considerably higher at BO3 than at the other investigated sea areas (Tukey's HSD-test; P<0.05). In addition, a signi®cant dierence in C:P and N:P ratios was found between F2 and LL12 (Tukey's HSD-test; P<0.05) and in N:P also between F2 and SR5 (Tukey's HSD-test; P<0.05).
Table 3 Carbon, nitrogen and phosphorus content (lg ind.±1) and the elemental ratios (lM:lM) of the dominant zooplankton species at the sampling stations (meanSD) (n number of replicates) Species
Station
Acartia sp.
SR5 IU5 LL3A LL12 LL23 BY15 LL23 BY15 F2
Centropages hamatus Daphnia cristata Eurytemora anis
Limnocalanus macrurus Temora longicornis
F2 SR5 IU5 LL3A LL12 F2 BO3 LL12 LL23 BY15
Carbon 2.90.3 1.10.07 2.70.2 1.30.5 2.10.3 1.70.2 3.20.4 2.70.3 1.50.2 2.40.5 1.60.2 1.70.1 2.30.3 1.70.2 31.32.9 34.94.3 2.10.7 1.80.4 2.00.2
n
Nitrogen
n
Phosphorus
n
C:N
C:P
N:P
4 3 3 3 3 3 3 3 3
0.60.08 0.20.02 0.50.03 0.20.1 0.40.04 0.30.04 0.70.1 0.60.09 0.30.05
4 3 3 3 3 3 3 3 3
0.100.02 0.050.01 0.100.005 0.080.009 0.060.01 0.060.004 0.120.007 0.120.007 0.070.008
3 3 3 3 3 3 3 3 2
5.60.2 5.80.3 6.20.1 6.30.3 6.00.1 5.70.2 5.10.04 5.30.2 6.30.4
82.527.8 59.820.7 70.66.1 40.817.6 96.728.3 77.28.4 66.49.0 60.111.1 58.812.4
14.95.1 10.53.9 11.41.1 6.63.2 16.24.4 13.61.0 13.01.8 11.42.5 9.42.4
3 3 3 3 3 3 5 2 3 3
0.50.1 0.30.05 0.30.02 0.40.08 0.30.05 6.90.7 6.70.6 0.40.2 0.40.07 0.40.06
3 3 3 3 3 3 5 2 3 3
0.090.005 0.060.007 0.070.009 0.070.006 0.080.01 0.790.04 0.630.09 0.080.02 0.100.02 0.070.01
3 3 3 3 3 3 5 2 3 3
5.80.4 5.80.2 5.80.3 6.20.5 6.10.4 5.20.1 6.10.2 5.90.3 5.60.4 5.70.1
72.011.1 68.65.8 43.93.3 89.83.9 57.72.3 102.611.6 143.820.6 63.06.8 45.32.6 71.116.0
12.52.3 11.80.7 7.60.6 14.51.7 9.50.4 19.42.3 23.52.9 10.81.8 8.10.2 12.52.9
429 Table 4 Statistically signi®cant between-site dierences (Tukey's HSD-test; P<0.05) in the mineral ratios (C:N, C:P, N:P) of the various zooplankton species (±, dierence insigni®cant). Location of sampling sites is shown in Fig. 1 (Aca Acartia sp.; Eur Eurytemora anis; Lim Limnocalanus macrurus; other taxa did not have any signi®cant dierences)
C:N F2 BO3 SR5 IU5 LL3A LL12 LL23 BY15 C:P F2 BO3 SR5 IU5 LL3A LL12 LL23 BY15 N:P F2 BO3 SR5 IU5 LL3A LL12 LL23 BY15
F2
BO3
SR5
IU5
LL3A
LL12
LL23
± ± ± ± ± ± ±
± ± ± ± ± ±
± Aca Aca ± ±
± ± ± ±
± ± ±
± Aca
±
Lim ± ± ± ± ± ±
± ± ± ± ± ±
± ± Aca ± ±
± ± Aca ±
Eur ± ±
Aca ±
±
± ± ± ± Eur ± ±
± ± ± ± ± ±
± ± Aca ± ±
Eur ± Aca ±
Eur ± ±
Aca Aca
±
Interspeci®c dierences Molar C:N ratio of all sampled zooplankton taxa ranged from 5.1 to 6.3. The C:N ratio diered signi®cantly between some of the zooplankton species (ANOVA; F5=5.6, P<0.0003); in the whole data the C:N ratio of C. hamatus was signi®cantly lower than that of Acartia sp., E. anis and Daphnia cristata (Tukey's HSD-test; P<0.05). Otherwise the C:N ratios of both the copepods and the cladoceran D. cristata were overlapping (Table 3). When samples collected at the same site were compared, there was a signi®cant dierence in the C:N ratio of Acartia sp. and C. hamatus at LL23 (the Baltic proper) and D. cristata and L. macrurus at F2 (the Bothnian Bay) (Tukey's HSD-test; P<0.05). The ranges of C:P and N:P ratios were 44±144 and 7±24, respectively. L. macrurus contained especially little phosphorus, and its C:P and N:P ratios were signi®cantly higher than those of the other species regardless of the sampling site (Tukey's HSD-test; P<0.05). Otherwise, the C:P and N:P ratios interspeci®cally overlapped due to the intraspeci®c variability (Table 3). Mineral composition of the zooplankton and the seston The stoichiometry of the zooplankton did not linearly follow the seston mineral ratios, and both positive and negative correlations occurred. There was a signi®cant negative correlation between the POC:PON ratio of the seston and the C:N ratio of Acartia sp. (Spearman's
correlation coecient (rs)=±0.7333; P<0.05; n=19), E. anis (rs=±0.487; P<0.05; n=15) and L. macrurus (rs=±0.845; P<0.05; n=8). The C:N ratio of C. hamatus correlated positively (rs=0.878; P<0.05; n=6) and its C:P ratio negatively (rs=±0.845; P<0.05; n=6) with the respective seston ratios. Further, the C:P and N:P ratios of L. macrurus had a signi®cant negative correlation with the respective ratios of the seston (Spearman's correlation matrix; P<0.05; n=8).
Discussion Mineral ratios The intraspeci®c C:N ratios of the dierent species were relatively stable, which conforms with stoichiometric theory (Sterner and Hessen 1994). The ratios were mostly overlapping at dierent sites, except for Acartia sp. and Limnocalanus macrurus. The C:P and N:P ratios of the zooplankton were more variable than the C:N ratio, as shown previously (Gismervik 1997a). Analysing phosphorus separately from the C:N measurements (which is common also in other studies), and combining the parameters in C:P and N:P ratios, may have aected the intraspeci®c variability, since the mean of the P replicates was used. Further, because the replicates were subsamples of a single net haul, the variability of the replicates could be underestimated. In spite of these limitations we see the mineral ratios as useful estimations and consider them comparable to other studies.
430
The intraspeci®c C:N, C:P and N:P ratios were generally in agreement with previous ®ndings for the corresponding season (Gismervik 1997a; Koski 1999). An exception was L. macrurus, since our C:N and C:P ratios were half the values determined earlier in the Baltic Fig. 2 Zooplankton mineral ratios in relation to the total seston mineral ratios (lM:lM) at the sampling sites (Aca Acartia sp.; Cen Centropages hamatus; Dap Daphnia cristata; Eur Eurytemora anis; Lim Limnocalanus macrurus; Tem Temora longicornis)
proper (Walve and Larsson 1999). This can be explained by the dierences in the sampling time, July versus September. Because L. macrurus stores lipid reserves in order to survive the winter, its carbon content and C:N and C:P ratios tend to be higher in the autumn than in
431
the summer. In contrast, the N:P ratios were comparable in both studies. Our results thus support the suggestion that the N:P ratio of the fat-storing zooplankton taxa neither depends on C content nor changes annually (Walve and Larsson 1999). The C:N ratios were also interspeci®cally overlapping. Only Centropages hamatus had a signi®cantly lower C:N ratio than Acartia sp., Eurytemora anis and Daphnia cristata, which is in accordance with the study of Gismervik (1997a). The C:P and N:P ratios of most of the species were close to each other, and the elemental ratios of the cladoceran D. cristata were also similar to those of the calanoid copepods studied. This is contrary to previous ®ndings (Andersen and Hessen 1991; Gismervik 1997a), because copepods are held to be relatively richer in nitrogen than cladocerans (Elser and Hasset 1994; Gismervik 1997a; Main et al. 1997). Instead, as also shown by Walve and Larsson (1999), L. macrurus had signi®cantly higher C:P and N:P ratios than all other species. These were due rather to the small phosphorus and higher nitrogen content of the species than to the high carbon content. Regardless of the changing seston elemental composition, zooplankton mineral ratio remained relatively steady (Fig. 2). Accordingly, there were no clear areal trends in zooplankton C:N, C:P and N:P ratios between the Baltic Sea basins. The high zooplankton C:N ratio in the Gulf of Finland, and the high C:P and N:P ratios in the Bothnian Bay and the Quark did not mirror seston, but were partly species related; C. hamatus, which was sampled only in the Baltic proper, had an especially low C:N ratio, and L. macrurus, collected only at the northern sites, was particularly poor in phosphorus. Mineral limitation of the zooplankton in the Baltic Sea According to stoichiometric theory, zooplankton maintains constant C:N and C:P ratios (Sterner and Hessen 1994), and can therefore become limited by food mineral content, if the food elemental ratios are high (Urabe and Watanabe 1992; Elser and Hasset 1994; Brett et al. 2000). The critical threshold for the food mineral ratio (Q*c-e) can be determined from Q*c-e=KcQz-e, where Kc is the gross growth eciency for carbon and Qz-e is the elemental ratio of the consumer (Urabe and Watanabe 1992). Kc can be assumed to be 30% (Straile 1997). Since the threshold estimate is based on the assumption that zooplankton do not excrete the limiting nutrient, the actual threshold is probably even lower than given by the equation (Urabe and Watanabe 1992; DeMott et al. 1998; Elser and Urabe 1999). If the critical threshold is lower than the elemental ratio of the food (C:N, C:P), grazers are potentially nutrient limited (N or P), while higher threshold ratios indicate C limitation and insuf®cient food quantity.
Our seston POC:POP ratios were higher than earlier measurements (100±110) in the Baltic proper (Walve and Larsson 1999) and the Gulf of Finland (113±191) in June and July (Koski and Viitasalo, unpublished data), whereas in lakes it is not unusual to have a seston C:P over 300 (Brett et al. 2000). Thus, our ratios still seem reasonable, since the Bothnian Bay, where POC:POP ratios up to 338 were measured, is nearly fresh water (salinity of 2 PSU), and primary production is typically P limited (Wul et al. 1986; Hansen 1996; present results), like in freshwater lakes. The Bothnian Bay has also a particularly high input of organic matter per unit volume (Wul et al. 1986), and a higher proportion of detritus (high POC:chl a) may have elevated the POC:PON and POC:POP ratios. Further, cyanobacteria can have a high C:P ratio (Larsson et al. 2001), which may have increased the seston POC:POP in the Gulf of Finland and the Baltic proper, where cyanobacterial aggregates covered the water surface. Nitrogen-®xing cyanobacteria also have relatively low C:N and high N:P ratios (Larsson et al. 2001). This could explain the lower seston POC:PON and the higher total seston PON:POP ratio in the Gulf of Finland and the Baltic proper in comparison to the Gulf of Bothnia. The seston POC:PON (>8) was always lower than the food threshold ratio (>17), and nitrogen should thus not limit the secondary production at any of the studied sites. Instead, grazer food may be phosphorus de®cient. D. cristata had the lowest threshold C:P ratio (196), followed by T. longicornis (19944), C. hamatus (21115), E. anis (22157), Acartia sp. (23864), and L. macrurus with a notably higher threshold ratio (41197). Critical thresholds indicated that all species, except L. macrurus, were potentially phosphorus limited in the Gulf of Bothnia, where the seston POC:POP was generally around 300. Some of the grazers were possibly P limited also in the Baltic proper and the Gulf of Finland (Table 5), but no phosphorus limitation of the zooplankton occurred in the eastern Gulf of Finland (LL3A) or the Quark (BO3). However, the results in the Quark were based solely on L. macrurus, which contained relatively little phosphorus. Therefore, it cannot be ruled out that P limitation occurred among other zooplankton species in the same area. The seston mineral ratios calculated here were generally much higher than some of the previous estimates for the critical food threshold ratio. If the threshold ratio (C:P) is 90±101 (DeMott et al. 1998) or 138±200 (Sterner and Hessen 1994), the Baltic Sea mesozooplankton could potentially be phosphorus limited. On the other hand, some authors suggest higher critical threshold ratios. A critical C:P ratio of 225±375 is possible for daphnids (Urabe et al. 1997; Brett et al. 2000), and, since copepods contain less phosphorus (Andersen and Hessen 1991; Gismervik 1997a; Hasset et al. 1997), their threshold ratio should be even higher. Accordingly, nutrient limitation of the zooplankton would be less obvious in the Baltic Sea, at least for L. macrurus.
432 Table 5 Potentially P-limited zooplankton species in the Baltic Sea, according to sampling station, during the study period. Estimates are based on calculation of the threshold elemental ratio of food (p P limitation; 0 no P or N limitation; ±, species not occurring
at the site). Threshold ratios did not indicate N limitation of any of the zooplankton species. The species are arranged according to their main area of occurrence, from north to south
Species
F2
BO3
SR5
IU5
LL3A
LL12
LL23
BY15
Daphnia cristata Limnocalanus macrurus Eurytemora anis Acartia sp. Temora longicornis Centropages hamatus
p 0 p ± ± ±
± 0 ± ± ± ±
± ± p p ± ±
± ± p p ± ±
± ± 0 0 ± ±
± ± p p p ±
± ± ± 0 p p
± ± ± 0 p p
As shown by Elser and Hasset (1994), another indicator of nutrient limitation is the elemental imbalance between seston and zooplankton, de®ned as: N:Pseston± N:Pzooplankton. They found that a positive imbalance, often observed in freshwater bodies, suggests phosphorus limitation of the zooplankton, while at marine sites a weaker or negative imbalance indicates nitrogen de®ciency. In the Baltic Sea, the seston N:P ratio nearly always exceeded the mesozooplankton N:P ratio (Fig. 3). Accordingly, the positive elemental imbalance N:Pseston±N:Pzooplankton suggested phosphorus limitation of the grazers. An exception to this was L. macrurus in the Quark (BO3), which showed a tendency to nitrogen limitation. However, the imbalance between N and P is relevant only if food quantity (carbon) is not limiting. Carbon limitation may occur at the times when seston nitrogen and phosphorus concentrations are high relative to carbon (Walve and Larsson 1999). Because the amounts of POC and chl a were lowest in the Gulf of Bothnia, especially the carbon-rich L. macrurus may suer from insucient food quantity rather than unsuitable combinations of the elemental nutrients. Because of the relatively low phytoplankton availability in the summer (Niemi 1975), it may also be harder for
Fig. 3 Elemental imbalance between <20 lm seston and zooplankton N:P ratios at the sampling sites. Values above zero indicate phosphorus limitation; values below zero indicate nitrogen limitation of the grazers. Abbreviations as in Fig. 2
the phosphorus-limited taxa to collect enough food particles to compensate for the P de®ciency of the food. Overall, both food availability and quality seem rather poor in the Baltic Sea in the summer. In the Gulf of Bothnia the seston was scarce and probably largely consisted of organic detritus (cf. Wul et al. 1986). In the Gulf of Finland and the Baltic proper, the seston was dominated by large-sized cyanobacterial aggregates, which are also held to be low-quality food (Koski et al. 1999; Brett et al. 2000). The seston and zooplankton samples were not, however, truly comparable in the Gulf of Finland and the Baltic proper. At these sites the zooplankton net was closed at 10±15 m depth to avoid taking up cyanobacteria in zooplankton samples, whereas seston was sampled in the surface waters. Still, the zooplankton samples should also contain migrating zooplankton, because sampling was carried out in the morning or early evening. The mineral composition of the food elucidates only one side of the zooplankton nutrition. The food quality is also modi®ed by several other properties such as cell size (Berggreen et al. 1988), cell morphology (Infante and Abella 1985; Burns 1987), toxicity (Lampert 1981, 1982; Nizan et al. 1986; Koski et al. 1999) and bio-
433
chemical compounds such as amino and fatty acids (Anderson and Hessen 1994; Gulati and DeMott 1997; MuÈller-Navarra et al. 2000). Brett et al. (2000) have suggested phosphorus limitation to determine the food quality for daphnids, while most other zooplankton species are more dependent on the highly unsaturated fatty acids (HUFA). Since phytoplankton taxa dier in their HUFA content, the grazers would be limited by fatty acids, if highly nutritious taxa like diatoms or cryptophytes are sparse in the phytoplankton community (Brett et al. 2000). Nevertheless, Gulati and DeMott (1997) have stated that the role of the fatty acids increases in importance only when the nutrient de®ciency of the grazers decreases. In the Baltic Sea, C or P limitation of the zooplankton may thus exceed the importance of HUFAs. Eect of zooplankton on nutrient cycles Because the primary production can be either P or N limited in the Baltic Sea, it is possible to compare whether the nutrient limitation of the Baltic zooplankton follows that of the primary producers. As in earlier studies (Wul et al. 1986; GraneÂli et al. 1990; Hansen 1996), the present water NO3:PO4 ratios indicated that P limitation of the primary production was most likely in the Gulf of Bothnia. This was not re¯ected in mesozooplankton, the mineral composition of which did not mirror seston elemental ratios. Instead, both D. cristata and the copepods were potentially limited by the concentration of particulate organic phosphorus, apart from in the eastern Gulf of Finland and the Quark. Our hypothesis of zooplankton acting as a phosphorus or nitrogen sink, depending on the phytoplankton nutrient limitation, was thus not fully supported. Because the Baltic zooplankton species seem to have a tendency towards stable mineral ratios, the grazers probably regulate their nutrient assimilation and excretion to maintain their elemental balance (Sterner et al. 1992; Gulati and DeMott 1997). To keep the balance in a phosphorus-depleted environment, Baltic Sea zooplankton should cycle P less eciently than N. The excess nitrogen is likely to be exported from the pelagic zone, if it is excreted in fast-sinking faecal pellets (Reigstad et al. 2000). However, in the Baltic Sea the bulk of the pellets are small and are probably remineralised in the upper water layers (Viitasalo et al. 1999). Therefore, zooplankton may signi®cantly aect the nutritional conditions of the water and hence the relationships between the dominating phytoplankton species in the Baltic Sea. Cyanobacterial blooms occur commonly in the Baltic Sea (Kononen et al. 1996). Since the nitrogen-binding cyanobacteria require a high supply of phosphorus (Smith 1983), an actively grazing zooplankton community may delay the initiation of the bloom by increasing the water N:P ratio. The ecient recycling of nitrogen could bene®t nitrogen-limited primary producers, since the regeneration of ammoni-
um by copepods favours phytoplankton growth (Touratier et al. 1999). Baltic Sea mesozooplankton have been suggested to enhance either N or P limitation of the autotrophs, depending on zooplankton taxa, gross growth eciency and season (Walve and Larsson 1999). The eects may, however, be minor due to small dierences in the stoichiometric ratios between the seston and the zooplankton (Walve and Larsson 1999). In the spring or early summer N limitation of secondary production may occur (Koski 1999; Walve and Larsson 1999), and both adult Acartia and Bosmina species can increase the N limitation of the primary producers. In contrast, during a higher seston N:P ratio in late summer, Acartia copepodites with a relatively low body N:P would tend to amplify P limitation (Walve and Larsson 1999). The present study suggests that also adult crustacean zooplankton may potentially accentuate P limitation of Baltic autotrophs. Hence, the Baltic Sea mesozooplankton resemble the grazers in freshwater bodies and fjords (Elser and Hasset 1994; Gismervik 1997b), acting as a phosphorus sink. Acknowledgements We wish to thank the Finnish Institute of Marine Research for the working facilities onboard R.V. ``Aranda'' and for the availability of salinity, nutrient and total chlorophyll a data; A. Nevalainen for conducting the C:N analyses; and E. Salminen and M. SjoÈblom for measuring the P contents at TvaÈrminne Zoological Station. This study was ®nanced by the Walter and Andree de Nottbeck Foundation (S.P.) and the Academy of Finland (M.K., M.V.). The shorelines in the map are based on coordinate data of the GEBCO Digital Atlas. The original source of the coordinate data is World Vector Shoreline, the US Defence Mapping Agency. The conducted experiments comply the laws of Finland.
References Andersen T, Hessen DO (1991) Carbon, nitrogen and phosphorus content of freshwater zooplankton. Limnol Oceanogr 36: 807±814 Anderson TR, Hessen DO (1994) Carbon or nitrogen limitation in marine copepods? J Plankton Res 16:317±331 Barrie A, Lemley M (1989) Automated 15N/13C analysis of biological materials. Int Lab 19:82±91 Berggreen U, Hansen B, Kiùrboe T (1988) Food size spectra, ingestion and growth of the copepod Acartia tonsa during development: implications for determination of copepod production. Mar Biol 99:341±352 Brett MT, MuÈller-Navarra D, Park SK (2000) Empirical analysis of the eect of phosphorus limitation on algal food quality for freshwater zooplankton. Limnol Oceanogr 45:1564±1575 Burns CW (1987) Insights into zooplankton±cyanobacteria interactions derived from enclosure studies. NZ J Mar Freshw Res 21:477±482 Cataletto B, Umani SF (1994) Seasonal variations in carbon and nitrogen content of Acartia clausi (Copepoda, Calanoida) in the Gulf of Trieste (northern Adriatic Sea). Hydrobiologia 292/ 293:283±288 Checkley Jr DM (1980) The egg production of a marine planktonic copepod in relation to its food supply: laboratory studies. Limnol Oceanogr 25:420±446 DeMott WR, Gulati RD, Siewertsen K (1998) Eects of phosphorus-de®cient diets on the carbon and phosphorus balance of Daphnia magna. Limnol Oceanogr 43:1147±1161
434 Elser JJ, Hasset RP (1994) A stoichiometric analysis of the zooplankton±phytoplankton interaction in marine and freshwater ecosystems. Nature 370:211±213 Elser JJ, Urabe J (1999) The stoichiometry of consumer-driven nutrient recycling: theory, observations and consequences. Ecology 80:735±751 Gismervik I (1997a) Stoichiometry of some marine planktonic crustaceans. J Plankton Res 19:279±285 Gismervik I (1997b) Implications of zooplankton stoichiometry on distribution of N and P among planktonic size fractions. J Plankton Res 19:343±356 GraneÂli E, WahlstroÈm K, Larsson U, GraneÂli W, Elmgren R (1990) Nutrient limitation and primary production in the Baltic Sea. Ambio 19:142±151 Grassho K, Ehrhardt M, Kremling K (eds) (1983) Methods of sea water analysis, 2nd edn. Verlag Chemie, Weinheim Gulati RD, DeMott WR (1997) The role of food quality for zooplankton: remarks on state-of-the-art, perspectives and priorities. Freshw Biol 38:753±768 Hansen HP (1996) Hydrochemistry. Nutrients. In: HELCOM (ed) Third periodic assessment of the state of the marine environment of the Baltic Sea, 1986±93; background document. Baltic Sea Environ Proc no. 64B, The Baltic Marine Environment Protection Commission, Helsinki, pp 212±214 Hasset RP, Cardinale B, Stabler LB, Elser JJ (1997) Ecological stoichiometry of N and P in pelagic ecosystems: comparison of lakes and oceans with emphasis on the zooplankton±phytoplankton interaction. Limnol Oceanogr 42:648±662 HELCOM (1996) Third periodic assessment of the state of the marine environment of the Baltic Sea, 1986±93; background document. Baltic Sea Environ Proc no. 64B, The Baltic Marine Environment Protection Commission, Helsinki Infante A, Abella SEB (1985) Inhibition of Daphnia by Oscillatoria in Lake Washington. Limnol Oceanogr 30:1046±1052 Kiùrboe T (1989) Phytoplankton growth rate and nitrogen content: implications for feeding and fecundity in a herbivorous copepod. Mar Ecol Prog Ser 55:229±234 Kononen K, Kuparinen J, MaÈkelaÈ K, Laanemets J, Pavelson J, NoÄmmann S (1996) Initiation of cyanobacterial blooms in a frontal region at the entrance to the Gulf of Finland, Baltic Sea. Limnol Oceanogr 41:98±112 Koski M (1999) Carbon:nitrogen ratios of Baltic Sea copepods ± indication of mineral limitation? J Plankton Res 21:1565±1573 Koski M, Klein Breteler WCM, Schogt N (1998) The eect of food quality on growth and development of the pelagic copepod Pseudocalanus elongatus (Copepoda: Calanoida). Mar Ecol Prog Ser 170:169±187 Koski M, EngstroÈm J, Viitasalo M (1999) Reproduction and survival of the calanoid copepod Eurytemora anis fed with toxic and non-toxic cyanobacteria. Mar Ecol Prog Ser 186:187±197 Lampert W (1981) Inhibitory and toxic eects of blue-green algae on Daphnia. Int Rev Gesamten Hydrobiol 66:285±298 Lampert W (1982) Further studies on the inhibitory eects of toxic blue-green Microcystis aeruginosa on the ®ltering rate of zooplankton. Arch Hydrobiol 95:207±220 Larsson U, Hajdu S, Walve J, Elmgren R (2001) Baltic Sea nitrogen ®xation estimated from the summer increase in upper mixed layer total nitrogen. Limnol Oceanogr 46:811±820 Main TM, Dobberfuhl DR, Elser JJ (1997) N:P stoichiometry and ontogeny of crustacean zooplankton: a test of the growth rate hypothesis. Limnol Oceanogr 42:1474±1478 Malej A, Faganelli J, Pedzic J (1993) Stable isotope and biochemical fractionation in the marine pelagic food chain: the jelly®sh Pelagia noctiluca and net zooplankton. Mar Biol 116:565±570
MuÈller-Navarra DC, Brett MT, Liston AM, Goldman CR (2000) A highly unsaturated fatty acid predicts carbon transfer between primary producers and consumers. Nature 403:74±77 Niemi AÊ (1975) Ecology of phytoplankton in the TvaÈrminne area, SW coast of Finland. II. Primary production and environmental conditions in the archipelago and the sea zone. Acta Bot Fenn 105:4±73 Nizan S, Dimentman C, Shilo M (1986) Acute toxic eects of cyanobacterium Microcystis aeruginosa on Daphnia magna. Limnol Oceanogr 31:497±502 Pagano M, Lucien SJ (1993) Organic matter, carbon, nitrogen and phosphorus contents of the mesozooplankton, mainly Acartia clausi, in a tropical brackish lagoon (Ebrie Lagoon, Ivory Coast). Int Rev Gesamten Hydrobiol 78:139±149 Reigstad M, Wassmann P, Ratkova T, Arashkevich E, Pasternak A, éygarden S (2000) Comparison of the springtime vertical export of biogenic matter in three northern Norwegian fjords. Mar Ecol Prog Ser 201:73±89 Smith VH (1983) Low nitrogen to phosphorus ratios favour dominance by blue-green algae in lake phytoplankton. Science 221:669±671 SoloÂrzano L, Sharp JH (1980) Determination of total dissolved phosphorus and particulate phosphorus in natural waters. Limnol Oceanogr 25:754±758 Sterner RW (1990) The ratio of nitrogen to phosphorus resupplied by herbivores: zooplankton and the algal competitive arena. Am Nat 136:209±229 Sterner RW (1993) Daphnia growth on varying quality of Scenedesmus: mineral limitation of zooplankton. Ecology 74: 2351±2360 Sterner RW, Hessen DO (1994) Algal nutrient limitation and the nutrition of aquatic herbivores. Annu Rev Ecol Syst 25:1±29 Sterner RW, Elser JJ, Hessen DO (1992) Stoichiometric relationships among producers, consumers and nutrient cycling in pelagic ecosystems. Biogeochemistry (Dordr) 17:49±67 Straile D (1997) Gross growth eciencies of protozoan and metazoan zooplankton and their dependence on food concentration, predator±prey weight ratio, and taxonomic group. Limnol Oceanogr 42:1375±1385 Touratier F, Legendre L, VeÂzina A (1999) Model of copepod growth in¯uenced by the food carbon:nitrogen ratio and concentration, under hypothesis of strict homeostasis. J Plankton Res 21:1111±1132 Urabe J (1993) N and P cycling coupled by grazers' activities: food quality and nutrient release by zooplankton. Ecology 74: 2337±2350 Urabe J, Watanabe Y (1992) Possibility of N and P limitation for planktonic cladocerans: an experimental test. Limnol Oceanogr 37:244±251 Urabe J, Clasen J, Sterner RW (1997) Phosphorus limitation of Daphnia growth: is it real? Limnol Oceanogr 42:1436±1443 Viitasalo M, Rosenberg M, Heiskanen A-S, Koski M (1999) Sedimentation of copepod fecal material in the coastal northern Baltic Sea: where did all the pellets go? Limnol Oceanogr 44:1388±1399 Voipio A (1981) The Baltic Sea. Elsevier Oceanogr Ser 30:1±418 Walve J, Larsson U (1999) Carbon, nitrogen and phosphorus stoichiometry of crustacean zooplankton in the Baltic Sea: implications for nutrient recycling. J Plankton Res 21: 2309±2321 Wul F, ártebjerg G, Nicolaus G, Niemi AÊ, Ciszewski P, Schulz S, Kaiser W (1986) The changing pelagic ecosystem of the Baltic Sea. Ophelia 4[Suppl]:299±319